a2 Georgetown University
The fixed effects estimator of panel models can be severely biased because of well-known incidental parameter problems. It is shown that this bias can be reduced in nonlinear dynamic panel models. We consider asymptotics where n and T grow at the same rate as an approximation that facilitates comparison of bias properties. Under these asymptotics, the bias-corrected estimators we propose are centered at the truth, whereas fixed effects estimators are not. We discuss several examples and provide Monte Carlo evidence for the small sample performance of our procedure.
(Online publication May 31 2011)
We are grateful for helpful comments by Gary Chamberlain, Shakeeb Khan, Whitney Newey, the editor, and three anonymous referees. We also benefited from comments by workshop participants at AUEB, Boston University, Columbia University, Federal Reserve Board, Georgetown University, Harvard/MIT, Johns Hopkins University, Iowa State University, Panel Conference sponsored by NSF-NBER-UCLA, Penn State University, Linz Time Series Workshop, Triangle Seminar, Université de Montréal, UC Irvine, UC San Diego, University of Cyprus, University of Pittsburgh, USC, and Yale University. We thank Ekaterini Kyriazidou for making her Gauss code available. The first author gratefully acknowledges financial support from NSF Grant SES-0921187 and SES-0819638. The second author gratefully acknowledges financial support from NSF Grant SES-0095132 and SES-0523186.